Artificial Intelligence
So, what is Artificial Intelligence?
First of all the word artificial means that it's something made by human beings and not naturally occuring. In our case a computer is the artificial thing we are working with.
Then intelligence means the ability to acquire and apply knowledge and skills.
Therefore artificial intelligence is the use of computers to acquire and apply knowledge and skills - just like a human being would. For example the use facial recognition technology to identify a person from a picture, video or live stream by comparing their face data with that stored in the facial recognition system's database. For example your phone unlocking when it 'sees' your face and refusing to open if it is not your face.
It's all made by humans, usually for use by humans. One thing you should note is that there is at least a human being using the system, even if just the developer himself/herself. Aha, I should start a #Humans4Humans Campaign by the way - and especially now that everybody is worried they'll be replaced by robots.
More examples
We are already using artificial intelligence in many different ways, especially those by tech giants like Google and Microsoft.
For instance, on your Android, you often use these features:
- Google Keyboard - Can predict and suggest what you might want to type next.
- Google Assistant - Can perform various tasks on voice command. E.g play music, text someone, search something online... etc.
- Google Lens - Can search for something on the internet using a photo from your gallery or camera. Can translate a foreign language etc.
- Chat GPT - Can write articles, code, summarize text, translate, do assignment, chat etc. Nowadays it's reading screenshots and creating images.
- The camera and Google Photos - Can detect faces, smiles, gestures and for some, even recognize people. Google's Photos app can group photos according to people. For example put together all photos where I am featured.
That's not all the applications that use A.I, but this are at the very basic, ones we use everyday.
Should you be afraid of A.I?
Yes and no.
Yes because, as is already happening, jobs are being replaced and the problem is that there is not as much effort being made in creating new jobs. Or perhaps, our resilience is yet to be pushed to the limit.
No because, even though jobs are being replaced, new jobs are emerging and even though not yet as prominently as they should be, new jobs will eventually emerge.
I work at a cyber cafe' helping people write resumes and job application letters. ChatGPT is able to write the same in a matter of seconds. So has my job really been replaced? Not entirely! Because instead of writing the cvs/resumes/letters myself, my new job now is to give ChatGPT the data and request it to generate the same. ChatGPT still needs me to provides the data. Also, a different job ChatGPT has created for me is showing people how to use it and use it optimumly. So I could soon find myself teaching and not typing. And I sure like teaching more than typing. So I am ready to embrace the future. You too should embrace technology for a better future instead of thinking about how you'll stick around with your good old job full of repetitive tasks that a computer programme can so easily replace.
A Developer's Perspective on Artificial Intelligence
I am also a software developer, or at least, have learnt to think like one.
I appreciate what new tecnology is doing. In fact, I love it and wonna be part of it. Since I learnt how to code, I have valued it to such the extent that I now take it to be the most useful thing I can do.
Artificial intelligence dawns on me as an application of computers with the highest potential of creating the biggest impact in every field it is applied. And so therefore, has drawn massive interest not only from me but also from many other people.
So, I wondered what exactly artificial intelligence is and why it's becoming more and more prominent now. And I realised that AI is more and more of specialized computer applications. For example, when it comes to facial recognition applications, lots of faces were studied, features were noted down, a face was defined and then methods of detecting a face or let's say faces were developed. This was then further developed to recognize people, recognize their gender, estimate their age, predict their mood and much more. e.g. GitHub Face Recognition API Repository
OpenCV also has it's own facial recognition technologies.
But let no one lie to you: Artificial intelligence is simply an imitation of human intelligence running on a computing device. You see all these amazing technologies perplexing you, they are simply someone's ideas autonomously running on a server, computer/smartphone or micro-controller (any computing device in short). Hopefully soon we will be able to develop apps that make it possible for non-programmers to build AI systems too.
Why is AI getting more and more prominet now. I realize there are two major reasons. First, more and more people as well as institutions and organisations have embraced technology demanding developers to create new bigger and better applications. Secondly, advancement in hardware. Better performing devices have allowed developers to create more sophisticated applications as well as enabling end-user devices (such as phones) to run more sophisticated applications. Also, the technique of splitting tasks among devices has enables us to run more powerful programs. For example, instead of doing everything using one supercomputer, ChatGPT uses many networked computers/servers.
The future with AI
We expect more and more tasks to be more automated.
It's up to demand (by clients or forced by necessity) and/or desires of developers and visionaries.
For example I came across this weeding robot, known as Tertill that uses solar power to patrol and automatically keep weeds out of your garden. This is an almost effectively and efficiently autonomous way of weeding as it doesn't require human intervention to operate. You don't even have to charge it - amazingly it runs on solar. This is very inspiring. It makes me desire to develop further on such a concept to also develop a robot to completely take over weeding jobs and other repetitive, time consuming jobs. If you live in a city you can think of cleaning robot - somehow related concept.
I have identified many other tasks I would like robots to take over. I once thought of an autonomous way of preparing breakfast. I know all these can be developed. In fact I want to be part of the team that develops them.
Well, I was born and live in Africa. See, my mum is a teacher but most of the time she is worried about her small farm and the animals on it (cows, sheep, goats and chicken) just like the other village women who are mostly subsistent small scale farmers. Not that this is wrong, but I have always thought she should be spending most of her time on better jobs that would improve other peoples' lives such as putting more time on her job as a teacher rather than doing farming in a traditional manner like the other village women. I am therefore very much interested in developing technologies to replace repetitive tasks such as weeding as well as improve the way we farm - to make it way easier and more productive.
It all starts with understanding what you want to achieve as a developer. Then learning and research, and finally development.
With technologies such as TensorFlow and availability of easily progammable hardware, more and more robots will be developed.
The Coming Robots...
I know you are imagining that an army of dangerous, human-structured, super-tech, walking robots are coming your way probably having been sent by a lab guy somewhere. Well that's the movie perspective. And it's as amusing as the movie guys want it to be.
But you are very wrong. As a developer, and hopefully other developers share my perspective too, no one is interested in building a massive walking piece of metal (or is it carbon fibre) for a robot, unless for amusement. Why, because it makes no sense. It's too much work trying to recreate a human being. And even if we were ready to put in the work and recreate one, it would never be perfect. Also in my view, human beings have 'serious weaknesses' I would not want to recreate. The very reason we are building robots is to help us in our various projects. So why would we put in all this effort to end up with these (human-like) robots with just the same weaknesses as ourselves? Weaknesses such as laziness, irrational decisions fueled by pride, greed etc.
The kind of robots coming are not those walking robots in your head. They are simply AI assisted machines. The very same machines we use in our daily lives, but now they will come with some intelligence. New machines will come up, but definately not walking robots.
For example, we have this large-scale weeding machine by Stout Industrial Technology that uses AI, object-recognition technology in particular, in combination with mechanics to identify and remove weeds.
Shall AI be more intelligent than humans
As defined earlier, an AI system is a specialized system. On the example of the weeding machine, the AI sytem is taught how a cabbage for example looks like, then it is programmed to recognize/identify any other plant that is not a cabbage as a weed. This way it can then take the action of weeding out all the plants that are not cabbages. It could even count them. As you can guess, counting all the weeds identified per view will only take micro-seconds if not nano-seconds. Counting the cabbages rather than the weeds makes more sense though.
So, does that mean that the AI system is more intelligent than a human being? No. The system is only doing math according to pre-defined rules. And you know computers are so good at math. So it's faster and more efficient than a human. It's performing specific instructions. . AI systems are also coded to learn and perform some instructions autonomously. For example, a self-driving car's system is taught to identify pedestrians and stop whenever a pedestrian is detected within a set range of distance. The pedestrian-detection system is also probably trained to learn from this data and predict places where there is likely to be a pedestrian. AI systems are therefore self-improving and that is why they keep out-performing humans.
Is AI more productive than humans?
The answer is yes. Well, AI is able to work faster and more efficiently for the task it's built for than humans.
But again, what's productivity? It's how effectively and efficiently specific goals or desired outputs are achieved using given amounts or resources including time.
So, if we put to task a weeding robot verses a human in a farm to weed, of course the robot would win - we don't even have to try. Man vs Machine! Whoa, machine wins!
When it comes to productivity, given that AI systems are specialized systems, AI systems are more productive. So no need to compete with computers doing what computers can do. Just let computers do it.
But then, who are these AI systems working for? Humans of course. Here is the fact: People using AI systems are more productive than those who do not.
It's more like a man using a tractor to plough vs a man using a jembe (hoe) to plough - and both given 10 acres to plough. Tractor man wins.
That means that we will see AI intergrate into almost every part of our lives.
Here is the opportunity
I have been very amused by a part of AI known as computer vision. Computer vision is the concept focussing on interpratation of images - from pictures, videos, live video streams i.e image data coming straight from your camera, etc. - and understanding them so as to extract meaningful information from them. Yeah, probably as a human being would.
Aha, like Google Lens does - can tell what's in a picture - a person, a cup, a tv, a puppy etc.
Facial recognition technologies also falls here. And many other things such as medical imaging, traffic flow monitoring, security survellance etc.
There are many beautiful things this technology can do.
Adobe Photoshop is already using this technology in photoediting. For example removing backgrounds, fixing another background, face-beauty etc.
It's the same technology being used in self-driving cars.
The beautiful thing is that it's being used to create visual aids for the visually-impared.
We need to develop technologies that solve our problems. For example we:
- Just came from a drought and need to develop technologies to combat the same. For example dig dams.
- Need to find a way to harvest and dry our maize lest it rots in the rain.
- Need to find alternative sources of energy as fuel prices and supply have become unpredictable.
AI and Development
Fortunately
Because of AI potential to increase productivity, it will spur and accelerate development sustainably.
Making work easier means more work can be done within the same amount of time - sometimes with less resources.
A farmer for example can spend time on activities that requires attention and specialization such as preparing feeds, studying the market etc. rather than doing weeding as it will be taken over by robots.
Unfortunately
In some places, people are yet to embrace technology.
I am a Kenyan. Kenya is a developing country as I guess you know. Our social structures have been very conservative. Innovation has simply not been part of our culture. I don't think there is someone who can stand and say that his grandfather invented something which went into production, and the family took the initiative to continue with improving the product and is today a brand! If your grandfather invented something, it is likely you don't know about it. And it is likely he really struggled to show the value of it. Making it to production was a dream, let alone creating a company which stands up to date. Our government policys and the way we think hasn't been supportive of innovation. That means that we have been slow in developing our own solutions. That's why we are not so much into manufacturing. We also have not stuructured ourselves to maximize on technology.
The sad bit is that we are yet to appreciate and support innovators/innovation to such a level as to grow our own campanies.
The saddest bit is that we combine capitalism and corruption in a state where politicians are the policy-makers and where most people are poor thereby putting them on the losing end.
We should realise that technology has the power to bring development - eradicating poverty and solving almost all of our problems (disease, inadequate housing, poor-education systems).
We should also realize that if we don't take action now, we will continue to depend on the west (Europe, America, Canada) and the West (China and Japan). Also, we shall continue to be caught in America vs Russia shows-of-might - for instance the current rise in prices of fuel as a result of sanctions and battles-for-control.
But recently, we have been improving - expanding our internet connections, teaching technology, even embracing electric vehicles & motorbikes.
How Lives Change
Just the same way we google-search almost everything or ask chatgpt almost everything, the same way we shall be asking our automated machines to do almost everything.
Aha, and that's where I fall in love with computer vision.
You see, the same way a self-driving cars recognizes other cars and knows how to drive itself, the same way any other machine will know how to work by itself. For example a farming system can check for weeds and schedule on itself a time to weed and proceed to do a clean weeding job on on its own.
Conclusion
Whatever can be done by a machine, the same can be automated, and whatever can be automated can be left for AI to do.
Remember AI is just an application of a computer. A computer does not have desires and goals of its own. A computer won't sit down and feel the desire to own a car, a house and marry a beautiful wife.
But a computer can help build that smart city and beautiful house you want to live in.
The robots are here. Ready and equiped for work. Asking "What can we do for you..?".
Written with passion for tech for better lives.
